On the Relationship between Development and Fertility: The Case of the United States On the Relationship between Development and Fertility: The Case of the United States Igor Ryabov Abstract: The present article addresses the question of whether there is a link be- tween the spatial patterns of human development and period fertility in the Unit- ed States at the county level. Using cross-sectional analyses of the relationship between Total Fertility Rate (TFR) and an array of human development indicators (pertaining to three components of the Human Development Index (HDI) – wealth, health, and education), this study sheds light on the relationship between fertility and human development. The analyses were conducted separately for urban, sub- urban and rural counties. According to the multivariate results, a negative associa- tion between selected human development indicators and TFR exists in suburban and rural counties, as well as in the United States as a whole. However, this is not the case for urban counties, where the results were inconclusive. Some indicators (e.g., median income per capita) were found to be positively, and some (e.g., the share of adults with at least bachelor’s degree) negatively, associated with TFR in urban counties. All in all, our results provide evidence of a negative relationship between human development indicators and period fertility in the United States at the county level, a fi nding which is consistent with the basic tenets of classic demo- graphic transition theory. Keywords: Period fertility · Development · Fertility-development relationship · United States · Second demographic transition 1 Introduction Recent research (e.g., Myrskylä et al. 2009) suggests that the negative relationship between human development and fertility in industrialized countries reverses when the Human Development Index (HDI) progresses beyond the threshold of 0.85-0.9. The present study seeks to further explore this relationship on the example of the United States. Some scholars (Lesthaeghe/Neidert 2006) regard the United States as an exception to the Second Demographic Transition (SDT) which is characterized by sub-replacement fertility, rising cohabitation and, more generally, the disconnec- Comparative Population Studies Vol. 40, 4 (2015): 465-488 (Date of release: 11.11.2015) © Federal Institute for Population Research 2015 URL: www.comparativepopulationstudies.de DOI: 10.12765/CPoS-2015-13en URN: urn:nbn:de:bib-cpos-2015-13en9 • Igor Ryabov466 tion of marriage and procreation (e.g., Lesthaeghe 2010; Lesthaeghe/Surkyn 1988; Van de Kaa 1987; Van de Kaa 1994). Using cross-sectional analyses of the relation- ship between period Total Fertility Rate (TFR) and selected human development in- dicators (life expectancy at birth for males and females, median income per capita, the share of population below the poverty line, the shares of adults without a high school diploma and with bachelor’s degree or higher), the present study explores the link between human fertility and development in the U.S. Our choice of human development indicators is based on an exploratory factor analysis which allowed us to identify the most relevant predictors of period fertil- ity at the county level. Each human development indicator corresponds to one of the three dimensions of the Human Development Index: wealth, health (longevity), and education (knowledge). HDI is generally the most commonly used index for hu- man development (Anand/Sen 1994; Noorbakhsh 1998). Not only has it been widely used for monitoring the relative development among all nations of the world, but it has also been applied to regions within countries (Acayaba/Oliveira 2013). Particu- larly, HDI has been used to quantify the regional differences in human development in the United States at the state and county levels (Burd-Sharps et al. 2008). This article examines the relationship between TFR and human development indicators by type (urban, suburban and rural) of county and county-equivalent ad- ministrative entities in the United States. The urban-suburban-rural classifi cation of counties used in the present study is based on the National Centre for Health Statistics (NCHS) classifi cation scheme, which is fundamentally a delineation of ge- ographical areas by urbanicity-rurality (Ingram/Franco 2012). Our classifi cation al- lows delineating urban, suburban and rural counties following the simplifi ed NCHS scheme (see section 6 for details). This article is organized as follows. The next two sections comprise a review of theoretical and empirical research devoted to the issue of socio-economic devel- opment and fertility. More specifi cally, the fi rst of these sections discusses fertil- ity regimes in advanced societies. A recent study by Myrskylä et al. 2009 is given special attention. The subsequent section addresses the applications of the Second Demographic Transition to the study of fertility regimes at the sub-national level, particularly in the United States. Next, we outline the rationale for the current study and the research hypotheses. A methodology section followsin which we describe data and variables employed in our empirical approach. In the results section which follows, we report descriptive statistics as well as multivariate analyses that model TFR as a function of human development indicators and racial-ethnic composition across the three types of U.S. counties (urban, suburban and rural). 2 Human Development and Fertility: Recent Evidence Research examining the relationship between human development and fertility has a long history (Luci/Thévenon 2010). Starting with Malthus (2013[1798]), according to whom fertility increases lead to the pauperization and moral degradation of large sectors of population due to the fi nite nature of natural resources, the weight of On the Relationship between Development and Fertility: The Case of the United States • 467 early scholarly opinion made a formulaic “truth” out of the negative relationship be- tween fertility and development (Myrskylä et al. 2009). While there exist a number of respectable theories predicting cyclical variations of fertility as a result of economic growth or decline (of which Malthusianism is a classical example), much scholarly attention has been dedicated to the demographic transition theory (DTT). Originally the theory was set to imply that in countries undergoing modernization, long-term economic growth should lead to a transition from high to low birth and death rates (Davis 1945; Notestein 1945). However, DDT was later expanded to predict ever-de- creasing fertility rates with economic growth in various kinds of contexts (Caldwell 1976, 1982; Kirk 1996; Lee 2003). Although in the second half of the 20th century DTT clearly dominated theoretical discussions related to fertility and development (Mason 1997), a more revisionist stance toward the issue was developed by the late 20th century. An infl uential report of the U.S. National Research Council (1989) pre- sented a view that departed, in important respects, from the classical interpretation of the fertility-development relationship offered by DDT. First, the report pointed out that population decline in and of itself can thwart economic growth. Secondly, and more importantly, it stressed that the correlation between slower population growth and decreased poverty does not imply causation. It is worth noting that when the report was published, the world’s TFR was approximately 3.4 children per woman (in contrast to today’s 2.3) and sub-replacement fertility was rare and limited to one region of the world – Europe (United Nations 2013). Although fertility below replacement level (defi ned as TFR<2.1) was a rarity 30 years ago, today it is widespread in Europe, Asia and the Americas. Fertility rates in some European countries has plummeted within the last dec- ade, thus creating the phenomenon of “lowest-low fertility”, defi ned as a total fertil- ity below 1.3 (Kohler et al. 2002). As of today, the lowest fertility levels are found in the Eastern European and Mediterranean countries (Billari/Kohler 2004; Kohler et al. 2002). At the same time, in highly developed countries, either the centuries-long fertility decline has stalled or an opposite trend has been observed simultaneously with continuing socio-economic development (Balbo et al. 2013; Bongaarts/Sobot- ka 2012; Goldstein et al. 2009). In north-western European countries (e.g., Denmark, Netherlands, and United Kingdom), where below-replacement fertility persisted at least for the length of one generation, there has been a fertility rebound since the late 1990s. Whereas contemplating the fact that fertility cannot decline ad infi nitum is quite compatible with DTT, explanations of the reversal of the fertility trend are not (Ma- son 1997). It is worth mentioning that one of the most prominent contemporary theories of post-transitional fertility, the Second Demographic Transition (reviewed in more detail below), predicts indefi nitely lower fertility in the context of industrial- ized countries as women obtain higher education and higher wages (Lesthaeghe/ Surkyn 1988). Paradoxically, however, fertility now appears to be rebounding in those countries which have progressed farthest on the path of the second demo- graphic transition (Goldstein et al. 2009). Moreover, these very countries are also the world’s leaders in gender equity – women’s labour market participation and tertiary education rates are the world’s highest, and still rising. Overall, we can sum- • Igor Ryabov468 marize that this brief overview of dominant theoretical approaches to reproductive behaviour shows shortcomings in explaining why long-term socio-economic devel- opment does not necessarily go hand-in-hand with a decline in period fertility rate. It has to be noted that, despite the long history of research on these questions, the fi eld has still not reached a consensus on the relative importance of different components of human development for fertility change (Bryant 2007; Lee 2003; Mc- Donald 2000). Traditionally, in human geography and across social sciences, much emphasis has been placed on economic determinants of fertility in the industrial- ized world (Bryant 2007). There are signs, however, that scholarly interest has be- gun to shift toward the social, non-economic components of human development (education, health, gender equity) as predictors of trends in period and cohort fertil- ity among developed countries (see, for example, studies of Goldstein et al. 2009; McDonald 2000; Myrskylä et al. 2012). A widely cited, and much discussed study by Myrskylä et al. (2009) empirically tested the hypothesis of an impact of the overall level of human development, prox- ied by the Human Development Index (HDI),1 on fertility, using cross-sectional and longitudinal data from the OECD area spanning from 1975 to 2005. In confronting cross-country data, the researchers contended that the development-fertility rela- tionship is negative when HDI levels are below the range of 0.85-0.9. However, when at the turn of the 21 century some countries reached an HDI above 0.9, the HDI- fertility association reverses to positive. Hence, by designating a clear turning point in the relationship between human development and fertility, Myrskylä et al. (2009) fi nd that human development is likely to induce a fertility rebound. It is reasonable to assume that the theoretical framework put forward by Myrskylä and colleagues can be used not only to explain cross-country variations in HDI-fertility association, but also to examine the effect of human development indicators on fertility within a particular country. The clear drawback of Myrskylä et al.’s (2009) work is that the effect of human development was represented by one integrated parameter, while its components (wealth, longevity and education) were not modelled as separate effects. Subsequent research (Furuoka 2009) argued that clarifying the effect of each component of the human development process is important for identifying the causal structure of the relationship between human development and fertility. 3 The Second Demographic Transition in the United States As mentioned above, the SDT is one of the dominant theories for explaining fam- ily formation and reproduction in the industrialized countries. Inspired by the rise of post-materialism (Inglehart 1971), the theory views marriage and reproduction as independent spheres and predicts the end of the predominance of the nuclear family as the only legitimate family type in the modern era (Van de Kaa 1987; Van 1 The human development index is a composite number that values between 0 and 1. A greater value of the HDI indicates a higher human development. On the Relationship between Development and Fertility: The Case of the United States • 469 de Kaa 1994). The key argument of the theory is that individuals in contemporary industrialized societies assign stronger importance to their own self-realization than to their family and children. According to this causal framework, the value shift toward individualism and self-realization is taken as an explanation for the delay of union formation and parenthood, an increased frequency of having several partners before the fi rst child, and a rise in cohabitation, which has been associated with a later age at entry marriage, if not a retreat from marriage. Ultimately, all these be- havioural changes lead to fertility decline (Lesthaeghe/Surkyn 1988). According to Lesthaeghe (2010), the new transition fi rst began in the Benelux countries in the 1970s and began to spread across Western Europe, Canada and Australia. North-western European countries have been among the fi rst to embrace a new family model by delaying childbearing and, simultaneously, a new economic model by investing in the educational and workplace opportunities of both men and women (Lesthaeghe 2010; Van de Kaa 1994). Furthermore, recent studies framed within the SDT theory argue that the second demographic transition would not re- main a regional idiosyncrasy, but would spread to the United States, Southern, Cen- tral, and Eastern Europe and industrialized Asian states (Lesthaeghe 2010; Raley 2001). Regardless of past demographic regimes in these countries, the SDT theory predicts the same demographic outcomes – the rise of divorce, cohabitation and out-of-wedlock childbearing – and most importantly – fertility decline. Of more direct relevance to this study is a paper by Lesthaeghe and Neidert (2006) which is devoted to the geographical pattern of the SDT in the United States. Examining geographical variations in the age at marriage, cohabitation rates and ex- tramarital fertility as correlates of the SDT, the authors found that the “blue states,” that is, the states that traditionally vote Democratic, exhibit an idiosyncratic pattern of the SDT, whereas the “red states,” that is, the states that traditionally vote Republi- can, exhibit traditional patterns of union formation. In other words, the conservative “red states” generally stick to the “breadwinner” model of the family that celebrates marriage as the institution ordained to promote the unity of sex, procreation and childrearing. In contrast, the “blue states” have moved toward the SDT by delaying childbearing and embracing a more liberal view of family. The authors conclude that the political divide in the U.S follows the same geographic pattern as does the SDT. There is one apparent weakness of Lesthaeghe and Neidert’s (2006) study – all analyses are conducted at the state level. Although many demographic policies are endorsed at the state level, it is worth mentioning that counties generally are the pri- mary political units of local government and have programmatic importance at the federal and state levels. It is also worth mentioning that many socio-demographic indicators vary signifi cantly at the county level. For example, TFR varies from 0.5 to 3.6. The percentage of adults holding bachelor’s or a higher educational degree varies from 4 to 70 percent at the county level, while the corresponding minimal and maximal values for the share of non-Hispanic whites are 1 and 100 percent. • Igor Ryabov470 4 Urban-Rural Divide and the Issue of Geographic Scale Changing geographic scale (the size of geographic units) can dramatically change estimated parameters. The United States is not only a vast country, but it is also known for its remarkable geographic fractionalization of demographic outcomes (Frey 1996; Massey and Denton 1988; Zelinsky 2011). Measuring American frac- tionalization calls for geographically precise data. For example, as our preliminary analyses indicate (not shown for parsimony), geographic clustering of fertility pat- terns rarely follows state lines. This is because prior waves of fertility change that had swept through the country produced over time a remarkable geographical pedigree of fertility regimes (Lesthaeghe and Neidert 2006). The “blue states” (e.g., California, New York) in Lesthaeghe and Neidert’s (2006) study contain, in addition to some of the largest urban areas in the country, a number of less urbanized and even rural areas that demographically are not distinguishable from similar areas of the neighbouring “red states”. In other words, examining geographic patterns of fertility behaviour at the level of U.S. states does not capture variations in fertility within states. No less important than the problem of geographic scale to this project is the issue of urban-rural disparities in fertility. According to Lesthaeghe and Neidert’s (2006), the TFR is higher in those states which are “more rural than metropolitan.” (p. 27). This fi nding points to the familiar pattern, by no means exclusive to the United States, of a rural “surplus” of fertility. Indeed, numerous studies across the globe (including the U.S.) demonstrate that, on average, urbanites tend to have less children than rural residents (Caldwell 1982; Gallup et al. 1999; Glenn/Hill 1977; Sharlin 1986). Moreover, considering human development indicators and fertility together, it becomes evident that there is an overlap between spatial patterns of human de- velopment and fertility. Unfortunately, much of what we know about urban-rural dimension in the relationship between human development and fertility is based on a small number of studies. However, there is an ample body of research doc- umenting the relationship between urbanization and development. Incorporating empirical evidence from around the world, socioeconomic status (SES) is shown to be positively correlated with urbanicity, with lower poverty and higher educa- tion and income levels found in urban than rural areas (Bradshaw 1987; Cochrane 1983; London/Smith 1988; Sicular et al. 2007). Cities have higher concentrations of human capital – skilled workers are paid higher wages, and this tendency has been rising over time (Glaeser/Redlick 2009; Rauch 1993). A recent study by Singh and Siahpush (2014) also shows that there are noticeable health differences across ur- banization levels in the U.S., with the life expectancy at birth being 2.0 years higher in urban than in rural areas. On the Relationship between Development and Fertility: The Case of the United States • 471 5 Study’s Rationale Unfortunately, there have been no empirical studies devoted to the development- fertility relationship in the context of the United States. To cover this gap, this pa- per investigates the impact of selected human development indicators on TFR by county type (urban, suburban and rural) in the United States. The conceptual basis for this study draws from the empirical evidence that the relationship between hu- man development and fertility in advanced industrialized societies is not necessar- ily negative, as is the case in developing nations (Goldstein et al. 2009; Myrskylä et al. 2009). Particularly, it has been argued that there could exist a threshold after which period fertility does not decline with increasing levels of socio-economic de- velopment (Myrskylä et al. 2009). Among the industrialized countries, the United States is characterized not only by relatively high human development, but also by high period fertility, especially in contrast to the sub-replacement fertility patterns of Europe (see Table 1). However, recent research confi rms that there are strong regional differences in correlates of SDT, with some states advancing further toward SDT and some not (Lesthaeghe/ Neidert 2006). The states that exhibit fertility and family change patterns which are more characteristic of SDT are usually more urbanized (Lesthaeghe/Neidert 2006). It has to be noted at the outset that the theories (i.e., DDT and SDT) that underlie this research have been constructed at the macro level for the entire world, and for regions to reveal the general trajectory of reproductive regimes. However, this investigation is based on the analyses conducted at the county level within one country. Thus, the general objective of the present paper is to test these theories at the county level. Particularly, this study will assess whether the relationship be- tween human development and fertility at the U.S. county level is indeed positive, as has been predicted by Myrskylä et al. (2009) for the industrialized countries. It is important to mention that the SDT theory predicts the opposite of what we expect to fi nd in this study (Lesthaeghe/Neidert 2006). Our main hypothesis is that each Tab. 1: TFR and HDI of Selected Industrialized Countries Source: CDC, Eurostat, UNDP Country USA France Germany UK TFR HDI TFR HDI TFR HDI TFR HDI 1980 1.8 0.825 2.0 0.722 1.6 0.739 1.9 0.735 1985 1.8 0.839 1.8 0.741 1.4 0.752 1.8 0.747 1990 2.1 0.858 1.8 0.779 1.5 0.782 1.9 0.768 1995 2.0 … 1.7 … 1.3 … 1.7 … 2000 2.1 0.883 1.9 0.848 1.4 0.854 1.6 0.863 2005 2.1 0.897 1.9 0.867 1.3 0.887 1.8 0.888 2010 1.9 0.908 2.0 0.879 1.4 0.904 1.9 0.895 • Igor Ryabov472 and every human development indicator has a positive impact on fertility. Thus, per each component of human development (wealth, health, and education), we expect that more affl uent, healthier and better educated U.S. counties will have higher fer- tility. We also predict that the urban-rural divide will have a moderating effect on the relationship between human development and fertility. Specifi cally, consistent with the premises of the SDT theory (e.g., Lesthaeghe/Neidert 2006; Van de Kaa 1994), period fertility is expected to be lower in urban counties than in rural counties, while suburban counties will occupy the middle ground between the two. This prediction is derived from the fact that urbanites were at the forefront of the SDT (Buzar et al. 2007). According Lesthaeghe and Neels (2002), urbanicity is one of the main struc- tural predictors of the SDT. 6 Methodology The study sample is drawn from several sources. The information on all independ- ent variables, except health, was obtained from the American Community Survey (ACS), arguably the most recent and complete source of offi cial statistics at the county level. The fertility data – Age Specifi c Fertility Rates (ASFR) by county – were obtained from the National Centre for Health Statistics (NCHS). The NCHS data cover the period of fi ve years (2007-2012) which allows aggregating the irregular and fragmented natality data from less populated counties. The dependent variable is Total period Fertility Rate (TFR), perhaps the most commonly used fertility indicator. Age Specifi c Fertility Rates (ASFR) were calcu- lated from the county-level NCHS natality data (vital statistics) averaged over the period of 2007-2012, while the data on female population by age were obtained from the ACS. TFR is averaged over the period of 5 years to adjust for relatively small numbers of births in certain age groups in less populated rural counties. In order to identify the most important human development indicators at the county level, we carried out several explorative analyses (not shown for parsimony). Specifi cally, per each component of the Human Development Index (wealth, health and education), we identifi ed between four and six theoretically relevant human de- velopment indicators obtained from the U.S. Census or vital statistics (via ACS and NCHS) and conducted an exploratory factor analysis. Per each indicator we selected two variables that loaded highly on one factor. To account for heteroskedasticity and autocorrelation, indicators that were highly correlated with other indicators were excluded, as were indicators for which there was little or no variation across counties. Thus, we obtained three factors that were interpreted by us as wealth, health and education. The variables that scored highly on health were (1) male and (2) female life expectancies at birth. The selected wealth indicators were (3) median income per capita and (4) the percentage of households with income below the poverty line. Finally, the two indicators that loaded highly on education were (5) the percentages of adults 25 years and older with less than a high school education and (6) with bachelor’s degree or higher. Importantly, we included an indicator that con- trols for the racial-ethnic composition of population – percentage of non-Hispanic On the Relationship between Development and Fertility: The Case of the United States • 473 whites – because fertility differs widely between racial-ethnic groups. It is also im- portant to note that racial-ethnic minorities score signifi cantly lower on some, if not all, human development indicators in the U.S. (Oliver/Shapiro 2006; Olshansky et al. 2012; Shapiro et al. 2013). Specifi cally, race gaps in longevity and education are a well-established fact (Guralnik et al. 1993; Olshansky et al. 2012; Ryabov 2011; Singh/Siahpush 2014). Thus, given the fi ndings of these studies, we reasoned that not controlling for racial-ethnic population composition would signifi cantly eschew our results. The distinction between urban, suburban and rural counties is based on the NCHS urban-rural classifi cation scheme, which was developed for monitoring the health of urban and rural residents (for details see Ingram/Franco 2012). The NCHS groups U.S. counties and county-equivalent entities into six urbanization levels (four metropolitan and two nonmetropolitan), on a continuum ranging from most urban to most rural. In short, counties are classifi ed as “large central metro” if they contain a Metropolitan Statistical Area (MSA) with 1 million or more inhabitants and satisfy at least one of three conditions: they (1) contain the entire population of the largest principal city of the MSA, or (2) have their entire population contained in the largest principal city of the MSA, or (3) contain at least 250,000 inhabitants of any princi- pal city of the MSA. According to the NCHS classifi cation, the next group – “large fringe metro” – includes those urban counties that belong to MSAs of 1 million or more inhabitants that did not qualify as “large central metro” counties. The next two types, “medium metro” and “small metro” counties, are differentiated by population size. “Medium metro” contains MSAs of populations from 250,000 to 999,999 and “small metro” MSAs have less than 250,000 inhabitants. Nonmetropolitan catego- ries are “micropolitan”, that is counties located in Micropolitan Statistical Areas, and “noncore” – rural counties that did not qualify as “micropolitan.” The detailed description of the NCHS classifi cation scheme can be found elsewhere (see Ingram/ Tab. 2: Types of the U.S. Counties by Urbanicity Urbanicity Description NCHS Classifi cation Urban Counties in MSAs* with 1 million or more residents Large central metro, large fringe metro Suburban Counties in MSAs* of populations of less than 1 million Medium metro, small metro Rural Counties containing no MSAs* Micropolitan,* noncore * In the United States, metropolitan and micropolitan statistical areas (metro and micro areas) are geographic entities delineated by the Offi ce of Management and Budget (OMB) for use by Federal statistical agencies in collecting, tabulating, and publishing Federal statistics. A metro area contains a core urban area of 50,000 or more popula- tion, and a micro area contains an urban core of at least 10,000 (but less than 50,000) population. Each metro or micro area consists of one or more counties and includes the counties containing the core urban area, as well as any adjacent counties that have a high degree of social and economic integration with the urban core. Source: Based on NCHS Classifi cation • Igor Ryabov474 Franco 2012). We simplifi ed the NCHS scheme by creating three urbanization levels (see Table 2). Our defi nition of urban counties includes “large central metro” and “large fringe metro.” The U.S. counties classifi ed in the NCHS scheme as “medium metro” and “small metro” are identifi ed by us as suburban. Finally, our defi nition of rural counties encompasses NCHS nonmetropolitan categories – “micropolitan” and “noncore.” 7 Results Table 3 gives the TFR values across the upper quartile, median, and lower quartile of each parameter (6 human development indicators plus the percentage of non- Hispanic whites) for a sample of 3,145 countries during the period 2007-2012. The bivariate analyses are also presented in the graphical form in Fig. 1-7. As shown in Table 3, academic achievement of Latino adolescents is associated with school SES and minority composition. As expected, academic achievement in schools in- creases with average school SES and decreases with the percentage of minority enrollment. As evident from Table 3 and Figures 1-7, an association between the selected human development indicators and TFR is likely, but the magnitude of the association varies depending on the variable in question. For example, the bivariate analyses indicate that a stronger association is likely between TFR and the percent- age of non-Hispanic whites than between TFR and life expectancy at birth (for both Tab. 3: Average TFR across Three Levels (High, Medium and Low)* of Human Development Indicators in the U.S. Counties (N=3,145; 2007-2012) Human Development Indicators Low Median High Wealth Median Income (dollars in 2010) 2.15 2.05 1.92 Poverty Rate (% below poverty threshold) 2.13 2.04 1.93 Health Male Life Expectancy at Birth (years) 2.07 2.04 2.00 Female Life Expectancy at Birth (years) 2.10 2.05 2.00 Education Less Than High School (%) 2.11 2.04 1.96 At Least Bachelor’s Degree (%) 2.13 2.05 1.94 Population Composition Non-Hispanic White Population (%) 2.19 2.05 1.89 * High, medium and low levels correspond to upper quartile, median and lower quartile of the variable distribution. Source: ACS, NCHS On the Relationship between Development and Fertility: The Case of the United States • 475 Fig. 1: Scatter Plot and exponential Trend Line of the Association between Median Income in 2010 and TFR in the U.S. Counties (N=3,145; 2007- 2012) $0 $10,000 $20,000 $30,000 $40,000 $50,000 $60,000 0 0.5 1 1.5 2 2.5 3 3.5 4 TFR Median Income Fig. 2: Scatter Plot and exponential Trend Line of the Association between Poverty Rate and TFR in the U.S. Counties (N=3,145; 2007-2012) Poverty Rate (% below federal poverty threshold) 0 10 20 30 40 50 60 0 0.5 1 1.5 2 2.5 3 3.5 4 TFR Source: ACS, NCHS • Igor Ryabov476 Fig. 3: Scatter Plot and exponential Trend Line of the Association between Male Life Expectancy at Birth and TFR in the U.S. Counties (N=3,145; 2007-2012) 70 72 74 76 78 80 82 84 0 0.5 1 1.5 2 2.5 3 3.5 4 TFR Male life expectancy (years) Fig. 4: Scatter Plot and exponential Trend Line of the Association between Female Life Expectancy at Birth and TFR in the U.S. Counties (N=3,145; 2007-2012) Female life expectancy (years) 72 74 76 78 80 82 84 86 0 0.5 1 1.5 2 2.5 3 3.5 4 TFR Source: ACS, NCHS On the Relationship between Development and Fertility: The Case of the United States • 477 Fig. 5: Scatter Plot and exponential Trend Line of the Association between Percentage of Adults with Less than High School and TFR in the U.S. Counties (N=3,145; 2007-2012) 0 10 20 30 40 50 60 0 0.5 1 1.5 2 2.5 3 3.5 4 TFR Fig. 6: Scatter Plot and exponential Trend Line of the Association between Percentage of Adults with at least Bachelor’s Degree and TFR in the U.S. Counties (N=3,145; 2007-2012) At least bachelor’s degree (%) 0 10 20 30 40 50 60 0 0.5 1 1.5 2 2.5 3 3.5 4 TFR Source: ACS, NCHS Less than high school (%) • Igor Ryabov478 women and men). Compared to TFR at the lowest quartile of male and female life expectancy (2.07 and 2.09, respectively), TFR corresponding to the lowest quartile of the distribution of percent non-Hispanic white is 2.19. Generally, the interquartile range in TFR is higher by at least a factor of two for the percentage of non-Hispanic whites and wealth indicators than for longevity indicators. Another noteworthy fact is that higher wealth, longevity and education correspond to lower levels of TFR, thus suggesting a pattern consistent with the DTT. The means of all variables by county type are shown in Table 4. The comparison of means between urban, suburban and rural counties reveals that there are quite a few urban-rural disparities in fertility and human development indicators. The fi rst noteworthy feature is that TFR is signifi cantly higher (p<0.01) in suburban and rural counties than in urban counties. Compared to suburban and rural counties, urban counties are characterized by signifi cantly higher incomes, lower poverty rates, higher female life expectancy, lower shares of adults without a high school degree and higher shares of adults with a college degree (bachelor’s or above). It is impor- tant to note that we also found signifi cant differences between suburban and rural counties in income and educational attainment. Additionally, urbanites tend to be more racially diverse than residents of suburban and rural counties. The share of non-Hispanic white population was 74.5 percent in rural counties, as compared to 53.7% in urban counties. Suburban counties occupied an intermediate position with respect to this parameter (64.0 percent). Fig. 7: Scatter Plot and exponential Trend Line of the Association between Percentage of Non-Hispanic White Population and TFR in the U.S. Counties (N=3,145; 2007-2012) 0 20 40 60 80 100 0 0.5 1 1.5 2 2.5 3 3.5 4 TFR White not latino population (%) Source: ACS, NCHS On the Relationship between Development and Fertility: The Case of the United States • 479 Multivariate results are presented in Tables 5 and 6. Table 5 models TFR as a function of seven predictors – median income per capita, population living in pov- erty (%), male life expectancy, female life expectancy, adult population with less than high school education (%), adult population with bachelor’s degree or higher (%) and non-Hispanic population (%) – in urban, suburban and rural counties, corre- spondingly. Each column in Table 5 shows the same regression model run on urban, suburban and rural counties. In contrast to Table 5, which shows the analyses con- ducted on three groups of counties, Table 6 presents an aggregate picture, predict- ing TFR in all U.S. counties. Multivariate regression model 1 in Table 6 is the same as the models in Table 5. Model 2 of Table 6 is the expansion of model 1. In addition to the aforementioned predictors, it includes the urbanicity level (reference: rural). Turning to Table 5 fi rst, observe that only one of the two effects monitoring wealth – income – is consistently signifi cant in all Table 5 models. Income is found to be conducive to higher fertility in urban counties (at p<0.01). At the same time, the association between income and TFR is negative in suburban and rural counties (at p<0.05). The effect for male longevity is negative and highly signifi cant, but only in rural counties. A negative association is also observed between life expectancy for females and period fertility in suburban and rural counties. However, longevity, re- gardless of gender differences, does not seem to matter when predicting variations Tab. 4: Means of Study Variables by U.S. County Type (N=3,145; 2007-2012) Study Variables U.S. Counties Urban Suburban Rural All (N=436) (N=731) (N=1,976) (N=3,143) Outcome Variables Total Fertility Rate 1.87ac 2.08a 2.15c 2.05 Independent Variables Wealth Median Income (2010 dollars) $34,974ac $28,980ab $25,383bc $29,103 Poverty Rate (% below poverty threshold) 14.6ac 15.4a 16.8c 15.4 Health Male Life Expectancy at Birth (years) 75.3 75.0 74.6 74.9 Female Life Expectancy at Birth (years) 81.8ac 80.3a 79.7c 80.1 Education Less Than High School (%) 12.2ac 14.3ab 16.8bc 14.5 At Least Bachelor’s Degree (%) 33.3ac 28.4ab 23.8bc 28.0 Population Composition Non-Hispanic White Population (%) 53.7ac 64.0ab 74.5bc 64.6 Note: Bonferroni method of comparing multiple means was used in the analysis. a Difference in the means signifi cant at p < 0.05 between urban and suburban counties. b Difference in the means signifi cant at p < 0.05 between suburban and rural counties. c Difference in the means signifi cant at p < 0.05 between urban and rural counties. Source: ACS, NCHS • Igor Ryabov480 in period fertility among urban counties. On educational attainment measures, U.S. counties seem to be split along the suburban/rural line. Only in rural counties is the percentage of adults who have not fi nished high school positively related to TFR. In urban and suburban counties, the other indicator of educational level is signifi - cant – that of highest level of attainment. Urban and suburban counties with lower percentages of adults with college degrees are predicted to have lower fertility. All in all, populations of those urban and suburban counties that are better educated are more likely to have lower period fertility. The effect of the share of non-Hispanic white population is uniformly negative across U.S. counties, regardless of their ur- banity level. Thus, U.S. counties with signifi cant shares of minority populations are likely to have higher fertility. Considered together, the fi ndings presented in Table 5 show that in rural and suburban counties, wealth, health and education are inversely Study Variables Urban Suburban Rural Counties Counties Counties (N=436) (N=731) (N=1,976) Wealth Median Income 0.16*** -0.12** -0.12*** (0.06) (0.05) (0.03) Poverty Rate -0.06 0.08 0.04 (0.05) (0.04) (0.03) Health Male Life Expectancy at Birth 0.07 0.05 -0.12*** (0.06) (0.04) (0.03) Female Life Expectancy at Birth 0.07 -0.11* -0.07* (0.05) (0.05) (0.03) Education Less Than High School 0.06 0.08 0.11*** (0.04) (0.04) (0.03) At Least Bachelor’s Degree -0.12*** -0.10* -0.05 (0.05) (0.05) (0.04) Population Composition Non-Hispanic White Population -0.15*** -0.23*** -0.14*** (0.05) (0.07) (0.04) Pearson’s R2 0.169 0.156 0.170 Tab. 5: Unstandardized Regression Coeffi cients and Their Standard Errors (in Parenthesis) in OLS Regression Models Predicting Total Fertility Rate in, Suburban and Rural Counties (N=3,145; 2007-2012) *p < 0.1; **p < 0.05; ***p < 0.01 Source: ACS, NCHS On the Relationship between Development and Fertility: The Case of the United States • 481 related to TFR. Whether this holds true for all U.S. counties or county equivalents is further investigated below (see Table 6). In model 1 of Table 6, there are four signifi cant (p<0.05) effects. Per capita in- come, female life expectancy at birth, percentage of college-educated, and percent- age of non-Hispanic whites are all negatively associated with TFR. Finally, the full model adds two dummies for urbanicity – urban and suburban (rural is the refer- ence category). While the effect for suburban is not signifi cant, the effect for ur- ban counties is signifi cant and negative. Consistent with earlier studies (e.g., Glenn/ Hill 1977; Lesthaeghe/Neidert 2006; Westoff 1954), this fi nding suggests that urban Tab. 6: Unstandardized Regression Coeffi cients and Their Standard Errors (in Parenthesis) in OLS Regression Models Predicting Total Fertility Rate in All U.S. Counties (N=3,145; 2007-2012) Study Variables Model 1 Model 2 Wealth Median Income -0.12** -0.12** (0.04) (0.04) Poverty Rate 0.08 0.08 (0.03) (0.03) Health Male Life Expectancy at Birth 0.05 0.05 (0.03) (0.03) Female Life Expectancy at Birth -0.11* -0.11* (0.03) (0.03) Education Less Than High School 0.08 0.08 (0.03) (0.03) At Least Bachelor’s Degree -0.12** -0.14*** (0.03) (0.03) Population Composition Non-Hispanic White Population -0.18*** -0.15*** (0.04) (0.04) Urbanicity (Reference: Rural) Urban -0.12*** (0.05) Suburban 0.08 (0.04) Pearson’s R2 0.171 0.174 *p < 0.1; **p < 0.05; ***p < 0.01 Source: ACS, NCHS • Igor Ryabov482 populations tend to have lower fertility. All in all, the results presented in Table 6 are quite similar to those presented in Table 5. Controlling for population racial-ethnic composition, the relationship between human development indicators and fertility is likely to be negative across suburban and rural counties as well as in the Unites States as a whole. 8 Discussion The determinants of fertility in industrialized societies are a recurrent topic in public policy debates on demographic policy (Balbo et al. 2013). One largely unsettled issue concerns the impact of human development on fertility in these countries. Al- though conventional wisdom posited a negative and resistant correlation between human development and fertility, recent evidence seems to contradict this asser- tion (Goldstein et al. 2009; Myrskylä et al. 2009). Myrskylä et al. (2009) argue that in advanced industrialized countries, further development halts the declining fertility rates. Another issue, often overlooked, is the impact of the place of residence on re- productive behaviour (Sharlin 1986). Urban–rural differences both in human devel- opment and fertility have long been recognized (Bradshaw 1987; Glenn/Hill 1977; London/Smith 1988). Specifi cally, it is largely assumed that urbanites tend to have smaller families than rural residents (Caldwell 1982; Glenn/Hill 1977; Sharlin 1986). In the context of the United States, strong regional contrasts have been found with respect to fertility and family formation patterns between more and less urbanized areas (Glenn/Hill 1977; Lesthaeghe/Neidert 2006). In the present paper, we addressed the question of whether there is an associa- tion between the spatial pattern of human development and period fertility in the United States at the county level. In order to address this question, we simultane- ously engaged with the issues of geographic scale, urbanicity-rurality and a choice of socio-economic factors that defi ne the concept of human development. Six hu- man development indicators – male life expectancy, female life expectancy, median income per capita, the share of population below poverty line, the shares of adults without high school diploma and with bachelor’s degree or higher – were selected as the ones representing the three domains of human development (wealth, health and education) and were regressed on TFR. The analyses were conducted sepa- rately for urban, suburban and rural counties. Following Myrskylä et al. 2009, we hypothesized that, holding all else constant, counties with elevated levels of human development would have higher levels of period fertility as well. In other words, we expected to fi nd a positive association between human development indicators and TFR. On the basis of literature documenting various degrees of association between spatial patterns of urbanicity-rurality and fertility (e.g., Cochrane 1983; Glenn/Hill 1977; Lesthaeghe/Neidert 2006; Sharlin 1986), we also expected that TFR should be lower in urban rather than rural counties. Our results reveal a pattern somewhat consistent across suburban and rural counties of a negative association between wealth, health and education, on the one hand, and period fertility, on the other. Hence, all things considered, our research On the Relationship between Development and Fertility: The Case of the United States • 483 hypothesis derived from Myrskylä et al. (2009) did not hold in the case of suburban and rural U.S. counties. Rather, our fi ndings provide evidence of support for the DTT and SDT theoretical models. The analyses carried out on all counties also showed that, as predicted at the outset, urban counties tend to have lower fertility. Urban setting is also unique in that only one human development indicator – the share of adults with at least bachelor’s degree – was found to have a negative effect on TFR in urban counties. Moreover, income in urban counties is found to be positively as- sociated with the outcome variable. In other words, the effects of wealth and educa- tion on period fertility in the urban United States seem to be in different directions. Apart from this, the effect of the ethnic composition of the population seems to be quite relevant in urban as well as rural settings. U.S. counties with lower percent- age of non-Hispanic whites tend to have lower period fertility, ceteris paribus. This result confi rms that there are clear differences in period fertility between the major racial-ethnic groups in the U.S. and these differences exist not only at the individual but also at the regional (county) level. It must be mentioned here that fertility dif- ferentials along racial and ethnic lines have existed in the United States for a long time and there have been identifi ed a number of analytical hypotheses concerning fertility differentials (Lichter et al. 2012; Parrado 2011). Although a full discussion of these hypotheses lies beyond the scope of the present study, we have to mention that racial and ethnic group differences in the distribution of social and economic resources are often credited as the main factor responsible for the persistent racial and ethnic fertility differentials in the U.S. (Parrado/Flippen 2012; Sweeney/Raley 2014). While we must be careful not to overstate the importance of these fi ndings given the limited explanatory power of our statistical models, it is worth noting that in U.S. urban areas the relationship between human development and fertility is not negative, as is likely to be the case in suburban and rural areas. It is also reasonable to assume that the trend toward low fertility in urban areas is not irreversible. Theo- retically, further increases in economic development could bring about a fertility rebound in U.S. urban areas. However, a growing share of college graduates can offset this rebound. It is also possible that, due to high costs of living in urban ar- eas, only high-income families can afford to stay in urban areas while their families are growing, while families of lesser means allocate to suburban areas. That would also explain why in urban areas the percentage of adults with at least a bachelor’s degree correlates with lower TFR while income does not. All in all, the analyses car- ried out in urban setting lend as much support to the SDT theory (e.g., Lesthaeghe/ Neidert 2006) as to the thesis put forward by Myrskylä et al. (2009). This research has limitations. To begin with, human development and fertility variations are highly interconnected. Although the present study assumes that pe- riod fertility is determined by socio-economic development, there are arguments in the literature concerning the ways in which fertility impacts socio-economic out- comes (e.g., Reher 2011). Secondly, period fertility measures like TFR are known to suffer from “tempo distortion” (Bongaarts/Feeney 1998; Bongaarts/Sobotka 2012). Although we do acknowledge that there is a discrepancy between period fertility measures and cohort experiences, we also need to point out the fact that intra- • Igor Ryabov484 regional differences in fertility are stable over time (Jones/Tertilt 2008; Pratt et al. 1984). For example, urban-rural differences in period fertility in the U.S. have been noted since the 1950s (Jones/Tertilt 2008; Westoff 1954).Thirdly, we did not account for the fact that some women can change residence during their pregnancy. Many rural areas do not have high-quality health personnel or facilities and, therefore, quite a few women residing in rural areas would prefer to travel to another, usually, suburban county for delivery and even for post-partum care (Starfi eld/Shi 2004). Although proximity to clinics that offer high-quality care is an important factor for pregnant women, anonymity is also an issue that is usually taken into consideration by adolescent mothers. Some women prefer to travel to more distant health care centres in order to ensure that their identity will not be revealed (Parkes et al. 2004). Therefore, it is possible that some births are registered in a county where the moth- er did not reside. Fourthly, we did not account for the modifi able areal unit problem (MAUP), a bias resulting from the arbitrariness of the geographical partition used. The essence of the MAUP is that there are many ways to draw boundaries to demar- cate space into discrete units, such as the U.S. counties, and the way boundaries are drawn conditions the choice of statistical approaches. Finally, we used only a few measures that are meant to capture the concept of “human development”. We can- not rule out the possibility that the results could have been different, had we used different human development indicators. From the perspective of future research, it is advisable to focus on contextual factors of regional differences in human development and fertility. Clarifying the contextual factors of the human development process is crucial for the investiga- tion of the geographic patterns of development and fertility. This applies no less to empirical investigation than to theoretical analysis. Further investigation is also warranted to assess spatial dependence and spatial autocorrelation in human de- velopment and fertility. According to our preliminary analyses (not shown), period fertility levels in nearby counties are more similar to each other than those in distant counties, regardless of urbanicity. Geographic Information System (GIS) models can address this problem by modeling spatial dependency. References Acayaba, Cíntia; Oliveira, Mariana 2013: Municipal HDI of Brazil grows 47,5% in 20 years, appoints UNDP (in Portuguese). Brasília, G1, Globo; July 29, 2013 [URL: http:// g1.globo.com/brasil/noticia/2013/07/idh-municipal-do-brasil-cresce-475-em-20-anos- aponta-pnud.html]. Anand, Sudhir; Sen, Amartya 1994: Human Development Index: methodology and mea- surement. No. HDOCPA-1994-02. Human Development Report Offi ce (HDRO), United Nations Development Programme (UNDP). Balbo, Nicoletta; Billari, Francesco C.; Mills, Melinda 2013: Fertility in advanced socie- ties: A review of research. In: European Journal of Population 29: 1-38 [doi: 10.1007/ s10680-012-9277-y]. On the Relationship between Development and Fertility: The Case of the United States • 485 Becker, Gary S.; Murphy, Kevin M.; Tamura, Robert 1994: Human capital, fertility, and economic growth. In: Becker, Gary S. (Ed.): Human Capital: A Theoretical and Empiri- cal Analysis with Special Reference to Education, 3rd edition. The University of Chi- cago Press: Chicago, IL: 323-350. Billari, Francesco; Kohler, Hans-Peter 2004: Patterns of low and lowest-low fertility in Europe. In: Population Studies 58,2: 161-176 [doi: 10.1080/0032472042000213695]. Bongaarts, John; Feeney, Griffi th 1998: On the quantum and tempo of fertility. In: Popu- lation and Development Review 24,2: 271-291. Bongaarts, John; Sobotka, Tomáš 2012: A demographic explanation for the re- cent rise in European fertility. In: Population and Development Review 38: 83-120 [doi: 10.1111/j.1728-4457.2012.00473.x]. Bradshaw, York W. 1987: Urbanization and underdevelopment: A global study of mod- ernization, urban bias, and economic dependency. In: American Sociological Review 52,2): 224-239. Bryant, John 2007: Theories of fertility decline and the evidence from Development Indicators. In: Population and Development Review 33,1: 101-127 [doi: 10.1111/j.1728- 4457.2007.00160.x]. Burd-Sharps, Sarah; Lewis, Kristen; Martins, Eduardo Borges 2008: The measure of America: American human development report, 2008-2009. Columbia University Press. Buzar, Stefan et al. 2007: Splintering urban populations: emergent landscapes of reurbanisation in four European cities. In: Urban Studies 44,4: 651-677 [doi: 10.1080/00420980601185544]. Caldwell, John C. 1976: Toward a restatement of demographic transition theory. In: Population and Development Review 2: 321-366. Caldwell, John C. 1982: Theory of Fertility Decline. London: Academic Press. Cochrane, Susan Hill 1983: Effects of education and urbanization on fertility. In: Bulatao, Rodolfo A. et al. (Eds.): Determinants of Fertility in Developing Countries, vol. 2: Fertil- ity Regulation and Institutional Infl uences. Academic Press: New York: 527-624. Davis, Kingsley 1945: The world demographic transition. In: The Annals of the American Academy of Political and Social Science 237: 1-11. Frey, William H. 1996: Immigration, domestic migration, and demographic balkaniza- tion in America: New evidence for the 1990s. In: Population and Development Review 22;4:741-763. Furuoka, Fumitaka 2009: Looking for a J-shaped development-fertility relationship: Do advances in development really reverse fertility declines. In: Economics Bulletin 29;4: 3067-3074. Gallup, John Luke; Sachs, Jeffrey D.; Mellinger Andrew D. 1999: Geography and eco- nomic development. In: International Regional Science Review 222: 179-232. Glaeser, Edward L.; Redlick, Charles 2009: Social capital and urban growth. In: Interna- tional Regional Science Review 35: 3-25 [doi: 10.1177/0160017609336079]. Glenn, Norval D.; Hill, Lester 1977: Rural-urban differences in attitudes and behavior in the United States. In: The Annals of the American Academy of Political and Social Science 429,1: 36-50. Goldstein, Joshua R.; Sobotka, Tomáš; Jasilioniene, Aiva 2009: The end of “lowest-low” fertility? In: Population and Development Review 35,4: 663-699 [doi: 10.1111/j.1728- 4457.2009.00304.x]. • Igor Ryabov486 Guralnik, Jack M. et al. 1993: Educational status and active life expectancy among older blacks and whites. In: New England Journal of Medicine 329,2: 110-116. Inglehart, Ronald 1971: The silent revolution in Europe: Intergenerational change in post-industrial societies. In: American Political Science Review 65,4: 991-1017. Ingram, Deborah D.; Franco, Sheila J. 2012: NCHS urban-rural classifi cation scheme for counties. In: Vital and Health Statistics, Series 2, Data Evaluation and Methods Research 154: 1-65. Jones, Larry E.; Tertilt, Michèle 2008: An economic history of fertility in the United States: 1926–1960. In: Rupert, Peter (Ed.): Frontiers of Family Economics, vol. 1. Em- erald Group Publishing Limited: 165-230. Kirk, Dudley 1996: Demographic transition theory. In: Population Studies 50,3: 361-387. Kohler, Hans-Peter; Billari, Francesco C.; Ortega, José Antonio 2002: The emergence of lowest-low fertility in Europe during the 1990s. In: Population and Development Review 28,4: 641-680. Lee, Ronald 2003: The demographic transition: three centuries of fundamen- tal change. In: The Journal of Economic Perspectives 17,4: 167-190 [doi: 10.1257/089533003772034943]. Lesthaeghe, Ron 2010: The unfolding story of the second demographic transi- tion. In: Population and Development Review 36,2: 211-251 [doi: 10.1111/j.1728- 4457.2010.00328.x]. Lesthaeghe, Ron; Neels, Karen 2002: From the fi rst to the second demographic transi- tion: An interpretation of the spatial continuity of demographic innovation in France, Belgium and Switzerland. In: European Journal of Population/Revue européenne de démographie 18,4: 325-360. Lesthaeghe, Ron J.; Neidert, Lisa 2006: The second demographic transition in the United States: Exception or textbook example? In: Population and Development Re- view 32,4: 669-698 [doi: 10.1111/j.1728-4457.2006.00146.x]. German translation: Der „Zweite Demographische Übergang“ in den USA: Ausnahme von der Regel oder Lehr- buchbeispiel?. In: Zeitschrift für Bevölkerungswissenschaft 32,3-4: 381-428 Lesthaeghe, Ron; Surkyn, Johan 1988: Cultural dynamics and economic theories of fer- tility change. In: Population and Development Review 14,1: 1-45. Lichter, Daniel T. et al. 2012: Hispanic assimilation and fertility in new US destinations. In: International Migration Review 46,4: 767-791 [doi: 10.1111/imre.12000]. London, Bruce; Smith, David A. 1988: Urban bias, dependence, and economic stagna- tion in noncore nations. In: American Sociological Review 53,3: 454-463. Luci, Angela; Thévenon, Olivier 2010: Does economic development drive the fertility rebound in OECD countries? Paper presented in the European Population Conference 2010 EPC2010, Vienna, Austria, September 1-4, 2010. Malthus, Thomas R. 2013[1798]: An Essay on the Principle of Population, Cosimo: New York. Mason, Karen Oppenheim 1997: Explaining fertility transitions. In: Demography 34,4: 443-454. Massey, Douglas S.; Denton, Nancy A. 1988: The dimensions of residential segregation. In: Social Forces 672: 281-315. McDonald, Peter 2000: Gender equity in theories of fertility transition. In: Population and Development Review 26,3: 427-439 [doi: 10.1111/j.1728-4457.2000.00427.x]. On the Relationship between Development and Fertility: The Case of the United States • 487 Myrskylä, Mikko; Kohler, Hans-Peter; Billari, Francesco C. 2009: Advances in devel- opment reverse fertility declines. In: Nature 4607(256): 741-743 [doi: 10.1038/na- ture08230]. Myrskylä, Mikko; Goldstein, Joshua R.; Cheng, Yen-Hsin Alice 2012: New cohort fertility forecasts for the developed world. MPIDR Working Paper WP 2012-014. Rostock: Max Planck Institute for Demographic Research. National Research Council 1989: Population Growth and Economic Development. Na- tional Academy Press: Washington, DC. Noorbakhsh, Farhad 1998: The human development index: some technical issues and alternative indices. In: Journal of International Development 10,5: 589-605. Notestein, Frank 1945: Population – the long view. In: Schultz, Theodore W. (Ed.): Food for the World. Chicago: University of Chicago Press: 36-57. Oliver, Melvin L.; Shapiro Thomas M. 2006: Black wealth, white wealth: a new perspec- tive on racial inequality. New York: Taylor and Francis. Olshansky, S. Jay et al. 2012: Differences in life expectancy due to race and educational differences are widening, and many may not catch up. In: Health Affairs 31,8: 1803- 1813 [doi: 10.1377/hlthaff.2011.0746]. Parkes, Alison; Wight, Daniel; Henderson, Marion 2004: Teenagers’ use of sexual health services: perceived need, knowledge and ability to access. In: Journal of Family Plan- ning and Reproductive Health Care 30,4: 217-224 [doi: 10.1783/0000000042177225]. Parrado, Emilio A. 2011: How high is Hispanic/Mexican fertility in the United States? Im- migration and tempo considerations. In: Demography 48,3: 1059-1080. Parrado, Emilio A.; Flippen, Chenoa A. 2012: Hispanic fertility, immigration, and race in the twenty-fi rst century. In: Race and Social Problems 4,1: 18-30 [doi: 10.1007/s12552- 012-9063-9]. Pratt, William F. et al. 1984: Understanding U.S. fertility: fi ndings from the National Sur- vey of Family Growth Cycle III. In: Population Bulletin 39,5: 3-42. Raley, R. Kelly 2001: Increasing fertility in cohabiting unions: Evidence for the second de- mographic transition in the United States? In: Demography 38,1: 59-66 [doi: 10.1353/ dem.2001.0008]. Rauch, James E. 1993: Does history matter only when it matters little? The case of city- industry location. In: The Quarterly Journal of Economics 108,3: 843-867. Reher, David S. 2011: Economic and social implications of the demographic transi- tion. In: Population and Development Review 37(s1): 11-33 [doi: 10.1111/j.1728- 4457.2011.00376.x]. Ryabov, Igor 2011: Adolescent academic outcomes in school context: Network ef- fects reexamined. In: Journal of Adolescence 34,5: 915-927 [doi: 10.1016/j.adoles- cence.2010.12.004]. Shapiro, Thomas; Meschede, Tatjana; Osoro, Sam 2013: The roots of the widening ra- cial wealth gap: Explaining the black-white economic divide. Institute on Assets and Social Policy: Research and Policy Brief, February 2013 [http://iasp.brandeis.edu. URL: http://iasp.brandeis.edu/pdfs/Author/shapiro-thomas-m/racialwealthgap- brief.pdf]. Sharlin, Allan 1986: Urban-rural differences in fertility in Europe during the demo- graphic transition. In: Coale, Ansley J.; Watkins, Susan C. (Eds.): The Decline of Fertil- ity in Europe: The Revised Proceedings of a Conference on the Princeton European Fertility Project: 234-60. • Igor Ryabov488 Sicular, Terry et al. 2007: The urban–rural income gap and inequality in China. In: Review of Income and Wealth 53,1: 93-126 [doi: 10.1111/j.1475-4991.2007.00219.x]. Singh, Gopal K.; Siahpush, Mohammad 2014: Widening rural–urban disparities in life expectancy, U.S., 1969–2009. In: American Journal of Preventive Medicine 46,2: 19-29 [doi: 10.1016/j.amepre.2013.10.017]. Starfi eld, Barbara; Shi, Leiyu 2004: The medical home, access to care, and insurance: a review of evidence. In: Pediatrics 113 (Supplement 4): 1493-1498. Sweeney, Megan M.; Raley R. Kelly 2014: Race, ethnicity, and the changing context of childbearing in the United States. In: Annual Review of Sociology 40: 539-558 [doi: 10.1146/annurev-soc-071913-043342]. United Nations 2013: World Population Prospects: The 2012 Revision. Highlights and Advance Tables, Working Paper No. ESA/P/WP. 228. United Nations Publications: New York. Van de Kaa, Dirk J. 1987: Europe’s second demographic transition. In: Population Bul- letin 42,1: 1-59. Van de Kaa, Dirk J. 1994: The second demographic transition revisited: theories and expectations. In: Beets GCN et al. (Eds): Population and Family in the Low Countries, 1993: Late Fertility and Other Current Issues, NIDI, CBGS. Amsterdam: Swets and Zeitlinger: 81-126. Westoff, Clyde F. 1954: Differential fertility in the United States: 1900 to 1952. In: Ameri- can Sociological Review 19,5: 549-561. Williams, David R.; Collins, Chiquita 1995: U.S. socioeconomic and racial differences in health: patterns and explanations. In: Annual Review of Sociology 21: 349-386. Zelinsky, Wilbur 2011: Not yet a Placeless Land: Tracking an Evolving American Geogra- phy. Amherst, MA: University of Massachusetts Press. Date of submission: 11.02.2015 Date of acceptance: 08.07.2015 Dr. Igor Ryabov (). The University of Texas – Pan American Department of Sociology. Edinburg. USA. E-Mail: ryabovi@utpa.edu URL: http://portal.utpa.edu/utpa_main/daa_home/cosbs_home/sociology_home/ faculty_staff Published by / Herausgegeben von Prof. Dr. Norbert F. Schneider Federal Institute for Population Research D-65180 Wiesbaden / Germany Managing Editor / Verantwortlicher Redakteur Frank Swiaczny Assistant Managing Editor / Stellvertretende Redakteurin Katrin Schiefer Copy Editor (German) / Lektorat (deutsch) Dr. Evelyn Grünheid Layout / Satz Beatriz Feiler-Fuchs E-mail: cpos@bib.bund.de Scientifi c Advisory Board / Wissenschaftlicher Beirat Paul Gans (Mannheim) Johannes Huinink (Bremen) Michaela Kreyenfeld (Rostock) Marc Luy (Wien) Clara H. Mulder (Groningen) Notburga Ott (Bochum) Peter Preisendörfer (Mainz) Zsolt Spéder (Budapest) Comparative Population Studies www.comparativepopulationstudies.de ISSN: 1869-8980 (Print) – 1869-8999 (Internet) Board of Reviewers / Gutachterbeirat Martin Abraham (Erlangen) Laura Bernardi (Lausanne) Hansjörg Bucher (Bonn) Claudia Diehl (Konstanz) Andreas Diekmann (Zürich) Gabriele Doblhammer-Reiter (Rostock) Jürgen Dorbritz (Wiesbaden) Anette Eva Fasang (Berlin) E.-Jürgen Flöthmann (Bielefeld) Alexia Fürnkranz-Prskawetz (Wien) Beat Fux (Salzburg) Joshua Goldstein (Berkeley) Karsten Hank (Köln) Sonja Haug (Regensburg) Hill Kulu (Liverpool) Aart C. Liefbroer (Den Haag) Kurt Lüscher (Konstanz) Emma Lundholm (Umeå) Nadja Milewski (Rostock) Dimiter Philipov (Wien) Roland Rau (Rostock) Tomáš Sobotka (Wien) Jeroen Spijker (Barcelona) Olivier Thévenon (Paris) Helga de Valk (Brussel) Heike Trappe (Rostock) Michael Wagner (Köln) © Federal Institute for Population Research 2015 – All rights reserved